Japanese 1989 rules (current Nihon Kiin official rules) say:

So after game end, if there is no agreement on the status of a group, there is a procedure whereby life or death of groups is determined. During this procedure, the only valid ko threat for a ko is a pass for that specific ko. (per section 7.2 quoted above)

I'm not even sure if there's much disagreement, it's just awkward to have a ruleset with these non-intuitive consequences.

Once I got to the point in my studies where I was used to leveraging different areas of the board against each other, I developed a profound dislike for the Japanese rules because of situations like this.

They're self-consistent, but just feel so kludgy. There's something lovely about a bent 4 being balanced by an appropriately sized unremovable ko threat.

_________________
Tactics yes, Tact no...

So if i understood correctly:

J1989: The game ends and there is no agreement on the status of a group. Now the life or death of groups is determinded, whilst doing so only passing is considered as an allowed ko-threat.

Therefore the bent four is dead because I don't have to fear the unremovable ko treat of the seki shape.

It would be unfair to thank only one person. I thank every single one in this thread for his time and contribution towards finding a solution.

But still: if i have to proof the group is dead i will lose 3 points (see my first answer), how's this handled?

No. All the stones put down in hypothetical play are removed from the board once the group in question has been proved to be dead or alive. (That's why it's called hypothetical.)

The life and death determination (if required) is just an evaluation step. Once the life or death has been determined, the board returns to the state after both players passed. In modern times, you can take a photo with a camera. Historically it had to be done on an adjacent board.

The Japanese 1989 rules have hypothetical play after the end of regular play to determine life and death if the players do not agree. Under hypothetical play the only thing that lifts a ban on taking a ko back is a pass in which the player designates the ko for which he is lifting the ban. In hypothetical play Black would not actually sacrifice the seki to make a ko threat. That was just to show the futility of doing so.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

The Japanese 1949 rules were criticized for having a number of ad hoc rules. The Japanese 1989 rules provided a rationale for nearly all of the ad hoc situations. (I cannot say that the '89 rules are logical, because I do not know of any computer program that applies them correctly. ) However, they introduced new anomalies, and the rules for hypothetical play are significantly different (non-intuitive) from the rules of regular play. I prefer the 1949 rules, myself.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

This weirdness about ignoring non-removable threats is completely separate from the concept of territory scoring, right? You could have a genuinely territory-based ruleset (not like AGA) that would treat bent four as seki when non-removable threats are present. All you need to do is get rid of that extra rule.

There are a number of territory scoring rules that do not involve hypothetical play to resolve life and death issues. Those that use an actual encore include Ikeda's rules, Lasker-Maas rules, and my rules. In all of these three the seki could be an unremovable ko threat. (Except for my specifically Japanese rules. )

BTW, some people believe that territory scoring is a relatively new aberration, that a form of area scoring came first. The evidence about which came first is unclear. The earliest known description of the rules appears to be for a form of area scoring. However, the earliest know game records that are scored appear to be for a form of territory scoring.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Herman, Japanese rules texts with precedental rulings or comments gave very specific examples. Two very similar shapes can have very different rulings. Also therefore, I do not overinterpret a bent-4 precedent. When there was an attempt to generalise a precedent, such an attempt was not welcome.

The Japanese 1989 Rules' pass-for-a-specific-ko rule cannot be applied literally because it creates contradictions to intentions. The Japanese 2003 Rules' generic-pass-for-ko rule should be applied for J1989 interpretation because it agrees to intentions whereever the J1989 rules writers were careful enough with their examples. For one bent-4, both ko-pass variants have the same behaviour.

http://home.snafu.de/jasiek/j1989c.html
http://home.snafu.de/jasiek/j2003.html

Xylol, in hypothetical analysis, a "pass" does not releave a [basic] ko ban. It requires the different (and artificial) move type "ko-pass" to releave a ko ban.

Proof-play is only imagined / hypothetical / executed with independent playing material / temporarily executed and then position and prisoners are restored. Therefore one does not lose points by playing to fill territory in proof-play.

Bill, it is not clear whether ko-pass is the only means of lifting a ko ban. It is (in theory) possible to have a ko, destroy it and resurrect it in the same local shape much later. Is it still the same ko? Undefined.

palapiku, territory scoring does not need special ko rules:

http://home.snafu.de/jasiek/sj.html

Thanks to everybody, you made the rules of the game I play a bit clearer